Simulation of draft force of winged share tillage tool using artificial neural network model

Abstract

An artificial neural network (ANN) model, with a back propagation learning algorithm, was developed to predict draft requirements of two winged share tillage tools in a loam soil. The input parameters to the 3–7–1 ANN model were; share width, working depth and operating speed. The output from the network was the draft requirement of each tillage tool. The developed model predicted the draft requirements of the winged share tillage tools with a mean relative error of less than 7% and mean square errors of less than 0.05, when compared to measured draft values. This result indicates that the ANN model had successfully learnt from the training data set to enable correct interpolation and could be used as an alternative tool for modeling soil-tool interaction under specific experimental conditions and soil types.